Antibiotic-Dispensing Practice in Community Pharmacies: Results of a Cross-Sectional Study in Italy

ABSTRACT

Inappropriate use of antibiotics in the community contributes to the development of antibiotic resistance (ABR), one of the most concerning issues in modern medicine. The objectives of the study were to investigate the knowledge and attitudes regarding ABR and dispensing antibiotics without prescription (DAwP) and to assess the extent of the practice of DAwP among Italian community pharmacists (CPs). A nationwide cross-sectional study using an anonymous, structured, validated, and pilot-tested questionnaire was conducted. The five sections gathered data on demographic and professional characteristics, knowledge and attitudes toward ABR and DAwP, practices regarding dispensing antibiotics with or without prescription and their reasons, counselling on the potential antibiotic side effects and the importance of adherence to medication regimen, and the information sources used to update the knowledge about ABR. About 4 in 10 CPs (37.1%) reported being involved in DAwP, although 93.7% knew that it is illegal in Italy. The vast majority affirmed to have always/often asked clients about their drug allergies (95.5%) and about their medication history (82.5%). Two-thirds (66.2%) warned their clients about the potential side effects of the drugs, and 55% informed them about the importance of completing the full course of antibiotics. Complacency with clients who found it difficult to consult the physician was the most significant predictor of DAwP. A considerable proportion of DAwP was described, so it could be easy for patients to misuse these drugs. Future policies need to enhance the enforcement of existing prescription-only regulations and to develop monitoring strategies to ensure their establishment in real-life practices.

INTRODUCTION

Antibiotic resistance (ABR) is one of the most concerning issues in modern medicine, due to its impact on morbidity, mortality, and socioeconomic factors (1–3). There are several reasons that contribute to the development of ABR, such as misuse and inappropriate use of antibiotics, especially in the community (4, 5). Several studies show how self-medication and over-the-counter (OTC) sales of antibiotics strongly influence the spread of ABR (6, 7). In this context, community pharmacists (CPs) could play a critical role in the struggle against ABR, as they are the professionals whom people address when they need drugs and antibiotics. Indeed, thanks to the strong relationship that CPs have established with the public, they are able to give information about ABR and the correct use of antibiotics (8, 43).

The practice of dispensing antibiotics without prescription (DAwP) enhances misuse and inappropriate use of antibiotics and, as a consequence, the development of ABR (9). In Italy, apart from some rare exceptions, DAwP is illegal, but despite this, the phenomenon is ever present and noteworthy (10, 11). According to the study “Antimicrobial resistance and causes of non-prudent use of antibiotics in human medicine in the EU” published in 2017 by the European Commission, 45.5% of the population in Europe has used antibiotics in the last 18 months and, among these, 8.7% have done so without a prescription (12). In Italy, 8.5% of people who have taken an antibiotic stated that they bought it without any prescription (12). The latest Eurobarometer survey showed that almost a third (32%) of European citizens have used antibiotics in the last 12 months and that 9% of Italian individuals have bought antibiotics without prescription (13). Understanding the causes of DAwP is the first step in designing effective interventions to control the phenomenon. A number of studies have explored factors which may influence the decision of CPs to dispense antibiotics without a prescription (14–16). A common trend with respect to factors associated with the practice of DAwP has been shown (9, 14–17). Factors intrinsic to pharmacy workers included insufficient knowledge, i.e., lack of continuing education (14), negative attitudes, i.e., complacency (9), fear of losing the client (14), and economic benefits (9, 15). Extrinsic factors included the client’s inability to afford a consultation with a physician (17), self-medication (14), and failure to enforce prescription-only use restriction policies (16).

The objectives of the present work were to describe knowledge and attitudes regarding ABR and DAwP and to assess the extent of the practice of DAwP among Italian CPs. A further aim of the study was to explore the factors potentially linked to DAwP.

RESULTS

Demographic and professional characteristics of respondents.

Of 500 CPs approached, 20 were deemed ineligible due to returned mail and/or no longer in business and 415 completed the survey, for a response rate of 86.5%. A total of 305 (73.5%) participants completed the survey online, and 110 (26.5%) did so through a telephone interview. More than half of the respondents (54.5%) were females, and the mean age was 50.1 (standard deviation [SD], ±9.9) years. Similarly, almost half of the Italian CPs (51.6%) were women, with a mean age of 56 years (18). About two-fifths of the pharmacies were located in metropolitan areas (39.8%). The CPs had a mean number of years in practice of 20.3 (SD, ±10 years), and almost two-thirds (64.1%) reported dispensing more than 200 medications per day. Only 11.1% of CPs reported previous work experience (6.5% as a medical sales representative and 5.1% as a hospital pharmacist).

Knowledge about ABR and DAwP.

The knowledge of the participants about ABR and DAwP is detailed in Table 1. Regarding the potential drivers of ABR, 78.1% and 67.2% of the CPs acknowledged early cessation of the antibiotic course and skipping one or two doses of antibiotics, respectively. Furthermore, 76.1% of the CPs were aware that ABR could lead to extension of hospital stay. The vast majority of the CPs participating in the study knew that DAwP is contributing to the spread of ABR (95.9%) and that it is illegal in Italy (93.7%). More than three-quarters of the respondents (78.3%) recognized that DAwP contributes to misuse of antibiotics by the public.

TABLE 1.

Respondents’ knowledge about ABR and DAwP^a

Statement	Response to statement
	Yes		No		Unsure
	n	%	n	%	n	%
ABR can lead to extension of hospital stay	316	76.1	60	14.5	39	9.4
Early cessation of an antibiotic course is one of the causes of ABR	324	78.1	70	16.9	21	5
Skipping one or more doses of an antibiotic contributes to the spread of ABR	279	67.2	89	21.5	47	11.3
DAwP contributes to misuse of antibiotics by clients	325	78.3	57	13.7	33	8
DAwP will lead to the spread of ABR	398	95.9	13	3.1	4	1
CPs can legally dispense antibiotics without a prescription in Italy	22	5.3	389	93.7	4	1

^aTotal number of observations, 415. ABR, antibiotic resistance; DAwP, dispensing antibiotics without prescription. The data for correct answers are shown in bold.

Attitudes toward ABR and DAwP.

Respondents’ attitudes toward ABR and DAwP are detailed in Table 2. Among positive attitudes, items ranged between 83.1% for the statement that the use of antibiotics without prescription can lead to inappropriate drug and dose choices (i.e., when a patient takes an antibiotic without an indication, such as for treatment of nonbacterial illness, or an incorrect dose, timing, or duration), to 97.1% for the belief that misuse of antibiotics can lead to ineffective treatment. In contrast, only 28.9% and 20.7%, respectively, strongly disagree or disagree that ABR is a problem that could be solved through the marketing of new antibiotics and that if the CP refuses to give antibiotics without prescription, the client can easily get them from another pharmacy, showing positive attitudes. Almost all believed that CPs should take an effective role in reducing ABR (96.1%) and that a client’s self-medication with antibiotics is one of the causes of ABR (94.2%).

TABLE 2.

Respondents’ attitudes toward ABR and DAwP^a

Attitude	Strongly agree or Agree		Uncertain		Strongly disagree or Disagree
	n	%	n	%	n	%
Client’s self-medication with antibiotics is one of the causes of ABR	391	94.2	18	4.3	6	1.5
Misuse of antibiotics can lead to ineffective treatment	403	97.1	8	1.9	4	1
ABR is a problem that could be solved through the marketing of new antibiotics	168	40.5	127	30.6	120	28.9
The occurrence of ABR is mainly a problem in hospital settings	155	37.3	117	28.2	143	34.5
If the CP refuses to give antibiotics without prescription, the client can easily get them from another pharmacy	238	57.4	91	21.9	86	20.7
CPs should take an effective role in reducing ABR	399	96.1	14	3.4	2	0.5
The use of antibiotics without prescription can lead to inappropriate drug and dose choices	345	83.1	56	13.5	14	3.4
The Ministry of Health should publish a clinical practical guideline on the dispensing of antibiotics and regulate such process	398	95.9	13	3.1	4	1
Antibiotics can be dispensed without prescription if a client finds it difficult to consult a physician	64	15.4	119	28.7	232	55.9

^aTotal number of observations, 415. The data for positive attitudes are shown in bold.

Practices regarding DAwP.

CPs were questioned about the frequency of their involvement in DAwP, and 37.1% reported to have been engaged in this practice. The most common conditions for which CPs dispensed antibiotics without prescription were toothache (82.5%), earache (45.5%), diarrheal diseases (34.4%), skin diseases (29.9%), and flu, common cold, and other upper respiratory tract diseases (22.7%). Antibiotics dispensed without a prescription were oral (81.8%) or topical forms (59.1%), eye drops (44.2%), ear drops (24.7%), and injectables (9.1%). The most commonly cited reasons for DAwP were the client’s promise to bring the prescription in the future (64.9%), personal knowledge of the client (44.2%), self-perceived knowledge about antibiotic use (31.8%), lack of client willingness to consult a physician for a nonsevere infection (24%), fear of losing a client (3.9%), and an increase in sales (2%).

Among 154 CPs who self-reported DAwP, the vast majority affirmed to have always or often encouraged the client to consult the physician (96.7%) and have always/often asked clients about their drug allergies (95.5%) and about their medication history (82.5%) (Table 3).

TABLE 3.

CPs’ practices regarding antibiotic dispensing

Statement	Never		Rarely		Occasionally		Often		Always
	n	%	n	%	n	%	n	%	n	%
When dispensing antibiotics (n = 415)
I warn clients about the potential side effects of the drugs	4	1	32	7.7	104	25.1	115	27.7	160	38.5
I advise clients about the importance of completing the full course of antibiotics	25	6	61	14.7	101	24.3	63	15.2	165	39.8
When dispensing antibiotics without prescription (n = 154)
I encourage clients to consult the physician	2	1.3	2	1.3	1	0.7	29	18.8	120	77.9
I ask clients if they are taking any other medication	0	0	10	6.5	17	11	22	14.3	105	68.2
I ask clients about drug allergies	0	0	0	0	7	4.5	20	13	127	82.5
I ask clients if they suffer from any liver and/or kidney disease	9	5.9	20	13	29	18.8	37	24	59	38.3

Results of multiple logistic regression analysis indicated that complacency with clients who found it difficult to consult the physician was the most significant predictor of DAwP (odds ratio [OR] = 0.47, 95% confidence interval [CI] = 0.31 to 0.71). Moreover, the odds of DAwP involvement increased with number of years in practice (OR = 1.03, 95% CI = 1.01 to 1.05) and was higher for females (OR =1.54, 95% CI = 1.01 to 2.36). Furthermore, having dispensed more than 200 medications per working day (OR = 0.58, 95% CI = 0.38 to 0.90) and working in a community pharmacy located in metropolitan areas (OR = 0.64, 95% CI = 0.42 to 0.99) showed an independent inverse association with DAwP (Table 4, model 1).

TABLE 4.

Multiple logistic regression analysis models

Model	Variable	OR	95% CI	P
1. Self-reported practice of DAwP (log likelihood = −260.034; χ² = 36.87; P < 0.0001; no. of observations = 415)	Belief that antibiotics can be dispensed without prescription due to inability to afford a consultation with a physician
	Yes	1.00a
	No	0.47	0.31–0.71	<0.001
	No. of yrs in practice	1.03	1.01–1.05	0.008
	No. of medications dispensed per day
	≤200	1.00a
	>200	0.58	0.38–0.90	0.014
	Location of community pharmacy
	Nonmetropolitan area	1.00a
	Metropolitan area	0.64	0.42–0.99	0.044
	Gender
	Male	1.00a
	Female	1.54	1.01–2.36	0.044
	Other work experiences
	No	1.00a
	Yes	0.55	0.27–1.12	0.098
	Knowledge score on ABR and DAwP	1.17	0.95–1.45	0.134
2. Having informed client about antibiotic use while engaged in dispensing (log likelihood = −240.374; χ² = 93.06; P < 0.0001; no. of observations = 415)	No. of medications dispensed per day
	≤200	1.00a
	>200	0.40	0.25–0.64	<0.001
	No. of yrs in practice	1.05	1.03–1.07	<0.001
	Belief that ABR is a problem that could be solved through the marketing of new antibiotics
	Yes	1.00a
	No	4.24	2.54–7.08	<0.001
	Knowledge score on ABR and DAwP	1.35	1.09–1.68	0.006
	Other work experiences
	No	1.00a
	Yes	0.51	0.24–1.05	0.068
	Belief that antibiotics can be dispensed without prescription due to inability to afford a consultation with a physician
	Yes	1.00a
	No	0.75	0.48–1.17	0.204
	Location of community pharmacy
	Nonmetropolitan area	1.00a
	Metropolitan area	0.84	0.53–1.33	0.452
	Gender
	Male	1.00a
	Female	1.10	0.71–1.71	0.674

^aReference category.

Regarding the CPs’ practices aimed at informing clients when dispensing antibiotics, two-thirds (66.2%) warned their clients about the potential side effects of the drugs, and 55% informed them about the importance of completing the full course of antibiotics (Table 3). The odds of having informed clients about antibiotic use while engaged in dispensing increased with number of years in practice (OR = 1.05, 95% CI = 1.03 to 1.07). Furthermore, having dispensed >200 medications per working day (OR = 0.4, 95% CI = 0.25 to 0.64) showed an inverse independent association with having informed clients while engaged in dispensing, whereas having a good knowledge score about ABR and DAwP (OR = 1.35, 95% CI = 1.09 to 1.68) and believing that the marketing of new antibiotics cannot solve the problem of ABR (OR = 4.24, 95% CI = 2.54 to 7.08) showed an independent positive association with the responsible practice of having informed clients while engaged in dispensing (Table 4, model 2).

Scientific journals were the main source of information about ABR (70.9%), followed by the Internet (61.7%) and continuing training courses (43.4%). However, more than three-quarters of the CPs (76.1%) reported an interest in more education on the topic of ABR.

DISCUSSION

As far as we know, the present investigation is the first attempt to estimate the extent of DAwP among Italian CPs. Moreover, given the urgent global need to address ABR, an understanding of the contributing factors for DAwP is helpful to identify possible solutions.

Knowledge about ABR and DAwP.

The majority of our sample showed appropriate knowledge toward ABR and DAwP. Although the vast majority of the CPs correctly answered the general questions about ABR and DAwP, around 22% of the participants did not know or were unsure that DAwP contributes to misuse of antibiotics by clients, and this finding is in line with a previous study performed in Hungary (19). Furthermore, we demonstrated that Italian CPs were not fully knowledgeable about the importance of completing the antibiotic course and avoidance of skipping one or more doses of an antibiotic. This result is of concern, since the lack of patient compliance with the antibiotic course is an important cause of ABR and the WHO Strategy for Containment of Antimicrobial Resistance encourages dispensers to inform clients on the proper use of antibiotics and the importance of completing the prescribed treatment (20). It is well established that taking antibiotics longer than necessary increases the risk of resistance, whereas the same effect from stopping antibiotic treatment early is not supported by strong evidence (21). This could partly explain the uncertainty of the CPs about this topic. As trusted health care professionals, CPs are in a unique position to help fight ABR by encouraging adherence to antibiotic regimens (22) and to instruct clients not to share or retain antibiotics for later use and to return any unused antibiotics to the pharmacy for disposal. However, ABR research and action should be reconfigured, taking seriously One Health as a mode of governance. Chandler has suggested that attention must be drawn to the configuration of ABR as a problem of connectedness through a One Health framework rather than approaches to implementation that most often target individual behavioral change (23).

Attitudes toward ABR and DAwP.

The finding that only one-third of the sample believed that ABR is not just a hospital setting problem is of concern, since numerous data show that resistant organisms have also been detected in patients in primary care (24). Most of third-generation cephalosporin-resistant Escherichia coli infections were acquired in the community, and only a minority (5%) occurred in the hospital setting (25). Moreover, new strains of methicillin-resistant Staphylococcus aureus (MRSA) emerged and rapidly spread throughout the community, causing severe infections among young and healthy individuals. These strains are infiltrating hospitals and replacing traditional hospital-acquired MRSA strains, with serious consequences for an elderly and debilitated patient population (26, 27). Although community-acquired resistant organisms are of particular concern, stewardship practices are much less prevalent in outpatient medicine than in hospital settings (28). Another important finding highlighted in the present study was the role of economic benefit, which could have an impact on DAwP. Indeed, more than half of CPs reported that if they refused to give antibiotics without prescription, the client could easily get them from another pharmacy, which is suggestive of fear of losing clients. It is well known that refusal to dispense antibiotics can affect sales and decrease profit margins, followed by the accumulation of stock of expensive medicines (15). In contrast, our study revealed that the role of patient-informer in reducing ABR was very much emphasized by the Italian CPs, since 96.1% believed that CPs should have a responsibility to take an effective role in reducing ABR. Public awareness campaigns with patient-pharmacist communication and involvement of pharmacists have been shown to be effective in several countries (29).

Practices regarding DAwP.

One of the most interesting findings of this study is that DAwP is common and practiced among Italian CPs, although outpatient antimicrobials are restricted by law to prescription-only use. About 4 in 10 CPs had dispensed antibiotics without prescription, so it could be easy for patients to misuse these drugs. Higher proportions were obtained in previous European studies with a similar methodology, which showed that antibiotics were supplied in 54.1% to 76.2% of cases (30, 31), whereas the percentage in the present study is worse than has been reported in Denmark (32) and Lithuania (33), where the proportions are 3.5% and 22%, respectively. The study also found that CPs with an increasing number of years in practice were more likely to dispense antibiotics without a prescription. Although not statistically significant, a similar trend was found in previous studies (9, 17). CPs with a higher number of years in practice might be the owners of the pharmacy and, therefore, could have a direct economic benefit by increasing sales. Moreover, it could be hypothesized that CPs with less work experience are more frequently involved in continuing education courses on ABR. The study also found that CPs who reported selling more than 200 medications per day and who worked in a metropolitan area were less likely to dispense antibiotics without a prescription. The difference in the proportion of DAwP could be attributed to the difference in CP response to client demand; it is possible that community pharmacies in a metropolitan area have higher income than those located in other areas and have no need to generate more revenues by DAwP. This is in accordance with the findings of previous studies which showed that variations in the prevalence rates of DAwP were linked to socioeconomic factors (34) and disparities in access to primary health care (17).

In most cases in which an antibiotic was dispensed without a prescription, the client was asked about his/her history of allergy and was encouraged to consult the physician. However, it is of concern that almost half of CPs never or occasionally advised clients about the importance of completing the full course of antibiotics. The latter result suggests that in an important number of cases, CPs did not provide the necessary information on the medications that are sold, as they assume that patients know the drugs they need. As expected, right attitudes, i.e., belief that ABR is a problem that could not be solved through the marketing of new antibiotics, and consistent knowledge about ABR and DAwP were predictors of having informed a client about correct use of antibiotics while engaged in dispensing.

Moreover, DAwP was also correlated with the belief that antibiotics can be dispensed without prescription due to the inability of the client to afford a consultation with a physician. Previous studies documented that CPs preferred to perform DAwP for minor health problems (35), whereas in clients with more severe clinical conditions, they believed the responsibility for prescribing antibiotics was restricted to physicians. The prescribing of antibiotics in these clinical situations calls attention to the need for educational interventions for CPs aimed at improving antibiotic use. However, the control of ABR cannot be the sole responsibility of health professionals. The public also has a major role to play. Public awareness should be encouraged to minimize patient demand for antibiotics and optimize appropriate use, especially in Italy, where self-medication with antibiotics for clinical conditions without indication (i.e., common cold and fever) has also been described (36).

Strengths and limitations.

The representativeness of the study population, obtained by drawing a sample stratified by geographical and metropolitan areas, represents a key strength of this survey. Moreover, this is the first survey conducted at a national level on this topic, and it acquires further importance since the DAwP is a driver of ABR and the impact of potential determinants is not fully known. Like most surveys, our study also has some limitations. There is a possibility of social desirability bias on the antibiotic-dispensing practice-related issues, where the respondents may give more favorable responses about their antibiotic-dispensing practice. This bias was limited by assuring participants that their responses would be anonymous and could not be traced back to them. Second, it is possible that people who were more informed or interested in the topic were more willing to participate and a nonresponse bias is attainable. However, the response rate, another strength in our study, was very high (86.5%). Third, the cross-sectional design of the study does not allow the drawing of conclusions on the direction of influence between the different variables and the outcomes of interest. However, this was not our primary goal, since we wanted to assess CPs’ knowledge and attitudes toward ABR and DAwP and practices related to DAwP. Finally, another possible limitation of this study, as is the case in all online surveys, is that the participants could look up the correct answers online or confer with someone before responding; nevertheless, the correct answer rate was similar to that of other studies (19, 35, 37, 38), so we are confident that this is not an issue in our data.

In conclusion, although it is forbidden by existing regulations, a considerable proportion of DAwP was described in the present study. Future policies need to enhance the enforcement of existing prescription-only regulations and to develop monitoring strategies to ensure their establishment in real-life practices.

MATERIALS AND METHODS

Study design.

We conducted a nationwide cross-sectional study in a representative sample of Italian CPs using an anonymous, structured, validated, and pilot-tested questionnaire. The first version of the questionnaire was pilot tested a month before the start of the study to evaluate the clarity of the items and to estimate the comprehensibility. The survey instrument was pretested using a cognitive interview with 40 CPs (20 via self-administered questionnaire and 20 via telephone interview) encouraging pretest respondents to voice their thought process while taking the survey. Findings from this phase of the study provided the research team with feedback that confirmed adequate comprehension of the questions, and minor refinements were made to improve flow.

The study was conducted between July 2019 and February 2020. The information was collected through a mixed technique, using an online questionnaire that was filled in by respondents and a direct telephone interview administered by an interviewer. We contacted CPs using the pharmacy’s telephone to inform CPs about the study and asked them to join the survey and whether they would prefer to answer a telephone interview or to complete an online survey. In the first page of the online survey, there was a personal data treatment information sheet at the end of which pharmacists could give their agreement to joining the study. Written or oral consent, the latter if a telephone interview was conducted, was obtained from the participants.

Study population.

To obtain a representative sample of Italian CPs, a multistage sampling design was used. First, the whole country was divided into three main geographical areas (north, central, and south and islands), and each area was subsequently divided into metropolitan and nonmetropolitan cities. Second, we randomly selected 8 municipalities belonging to the metropolitan centers and 60 municipalities not belonging to the metropolitan areas; thereafter, we randomly selected the same number of CPs from each cluster from a publicly available frame (39). The geographical subdivision used in our study (north, central, and south and islands) was according to the geographical division of the National Institute of Statistics (ISTAT) (40), which reflects the different characteristics of the resident population from a historical, economic, and cultural point of view. The choice of 8 metropolitan and 60 nonmetropolitan areas was driven by the Italian population distribution in 1,200 metropolitan and 7,000 nonmetropolitan municipalities.

Questionnaire.

The survey was developed after an extensive literature review (9, 14–17, 19, 35, 37, 38) to identify topics to cover in the study and to track opinions of Italian CPs about ABR and DAwP useful for comparisons with previous published studies. The final version consisted of 34 items divided into five sections. The first section (7 items, closed-ended items with multiple answers and open-ended items) gathered data on demographic and professional characteristics, including age, gender, number of years in practice as a CP, number of medications dispensed per day, and other work experiences (e.g., hospital pharmacist, medical sales representatives, etc.). The second section (6 items with a “yes,” “no,” or “unsure” response format) evaluated pharmacists’ knowledge about ABR and DAwP and its impact on the emergence of ABR and on public health. An overall knowledge score for ABR and DAwP was calculated by assigning one point for each correct response and summing the scores to each statement; the scores ranged from 0 to 6. The third section (9 items on a five-point Likert scale response format) evaluated CPs’ attitudes toward DAwP and ABR. The fourth section (10 items, closed-ended items with multiple answers and open option, and five-point Likert scale response format) evaluated CPs’ practices regarding dispensing antibiotics with or without prescription, including reasons for DAwP, and if they usually asked information about clinical history, warned clients about the potential side effects of antibiotics, and informed clients about the importance of completing the full course of antibiotics. The fifth section (2 items, closed-ended items with multiple answers and open option) explored the information sources that CPs used to update their knowledge about ABR and if they needed further information about it.

Ethical approval was granted by the Local Human Research Ethics Committee (identifier no. 123/2019/04/18).

Statistical analysis.

Statistical analysis was developed using STATA software program, version 16.1 (41).

Data were summarized using frequencies for categorical data and mean and standard deviations for continuous data. Multiple logistic regression models were developed according to the Hosmer and Lemeshow strategy (42), and independent variables for which the P value was 0.25 or less at the univariate analysis were included in the models. Furthermore, sociodemographic variables, regardless of the results of univariate analysis, were judged to potentially have influence on the outcomes of interest and were also included in the models. Bivariate and multiple logistic regression analyses were performed to determine the independent association of explanatory variables with the following outcomes of interest: having ever been involved in DAwP (model 1); having informed clients about antibiotic use while engaged in dispensing (model 2). In model 2, the outcome was identified by having warned clients about the potential side effects of the drug and having informed patients about the importance of completing the full course of antibiotics versus all others. The following explanatory variables were included in both models: gender (male = 0; female = 1), location of community pharmacy (nonmetropolitan area = 0; metropolitan area = 1), number of years in practice (continuous), number of medications dispensed per day (≤200 = 0; >200 = 1), other work experiences (no = 0; yes = 1), knowledge score about ABR and DAwP (continuous), and belief that antibiotics can be dispensed without prescription due to inability to afford a consultation with a physician (yes = 0; no = 1). In model 2, belief that ABR is a problem that could be solved through the marketing of new antibiotics (yes = 0; no = 1) was also included.

A P value of less than 0.05 was considered statistically significant. Adjusted odds ratio and 95% confidence intervals were calculated.

Data availability.

The data presented in this study are openly available in the Mendeley Data repository (https://data.mendeley.com/datasets/kg2cpcjg39/1).